AI Engineer Paris 2025 (Day 2)

The transcript captures the AI Engineer Paris event, featuring multiple speakers and discussions on AI engineering topics. The event includes presentations from industry leaders on subjects such as managing state in AI, the evolution of AI infrastructure, and advancements in generative AI models. Speakers share insights, demo new technologies, and engage in Q&A sessions, providing a comprehensive overview of the current state and future directions of AI engineering. The discussions cover a range of topics, from the challenges of building AI agents to the potential of open-source models and the importance of ethical considerations in AI development.

Outlines

Part 1: Introduction and State Management

Part 2: MCP and Agent Standardization

Part 3: Infrastructure and Scaling

Part 4: Image Generation and Editing

Part 5: MCP Deep Dive

Part 6: Agentic AI Stack and Infrastructure (Repetition)

Part 7: Flux Models (Repetition)

Part 8: Computer User Agents

Part 9: Knowledge Graphs and Multi-Agent Systems

Part 10: Open LLMs and Prompt Learning

Part 11: High-Performance Inference

Part 12: LlamaIndex and NotebookLM

Part 13: Google DeepMind and Generative Models

Part 14: Real-Time Voice AI

Part 15: LlamaIndex and NotebookLM (Repetition)

Part 16: Google DeepMind and Generative Models (Repetition)

Part 17: Real-Time Voice AI (Repetition)

Sign in to continue reading, translating and more.

Open full episode in Podwise

AI Engineer

Part 1: Introduction and State Management

13:33Introduction to AI Engineering and Managing State

Introduction to AI Engineering and Managing State

22:31A Brief History of Application Architecture and the Data Layer

A Brief History of Application Architecture and the Data Layer

28:12Four Properties of a Kick-Ass Data Layer for AI Applications

Four Properties of a Kick-Ass Data Layer for AI Applications

37:16Disambiguating Data and Neo4j's Vision

Disambiguating Data and Neo4j's Vision

42:25Neo4j Demo and Q&A: Agentic Memory and Innovation

Neo4j Demo and Q&A: Agentic Memory and Innovation

Part 2: MCP and Agent Standardization

52:59Docker and the Future of Agents

Docker and the Future of Agents

1:00:30MCP Protocol and Lessons Learned

MCP Protocol and Lessons Learned

1:11:11Challenges and Solutions in MCP Deployment

Challenges and Solutions in MCP Deployment

1:21:44The Future of MCP and Advice for Students

The Future of MCP and Advice for Students

Part 3: Infrastructure and Scaling

1:28:25Introduction to Coeb and the Agentic AI Stack

Introduction to Coeb and the Agentic AI Stack

1:37:41Defining AI Agents and Infrastructure Requirements

Defining AI Agents and Infrastructure Requirements

1:45:00Addressing Bottlenecks and Scaling Agentic Workloads

Addressing Bottlenecks and Scaling Agentic Workloads

1:51:21Q&A: Evolution of Infrastructure and Future Predictions

Q&A: Evolution of Infrastructure and Future Predictions

Part 4: Image Generation and Editing

2:00:16Unifying Text-to-Image Generation and Image Editing

Unifying Text-to-Image Generation and Image Editing

2:06:15Latent Flow Matching and Model Architecture

Latent Flow Matching and Model Architecture

2:11:11Model Speed and Community Engagement

Model Speed and Community Engagement

Part 5: MCP Deep Dive

2:21:28MCPs and Remote MCP Servers

MCPs and Remote MCP Servers

2:27:23Remote MCP Servers and Authentication

Remote MCP Servers and Authentication

2:33:34MCP Registries and Community Involvement

MCP Registries and Community Involvement

Part 6: Agentic AI Stack and Infrastructure (Repetition)

2:40:06Coeb and the Agentic AI Stack

Coeb and the Agentic AI Stack

2:45:00Defining AI Agents and Infrastructure Requirements

Defining AI Agents and Infrastructure Requirements

Part 7: Flux Models (Repetition)

2:50:22Black Forest Labs and Flux Models

Black Forest Labs and Flux Models

2:57:10Latent Flow Matching and Model Architecture

Latent Flow Matching and Model Architecture

3:03:31Model Speed and Community Engagement

Model Speed and Community Engagement

Part 8: Computer User Agents

3:11:14The H Company and Computer User Agents

The H Company and Computer User Agents

3:17:17Architecture and Efficiency of Computer User Agents

Architecture and Efficiency of Computer User Agents

3:22:31Training and Open-Source Philosophy

Training and Open-Source Philosophy

3:28:21Future Directions and Community Engagement

Future Directions and Community Engagement

Part 9: Knowledge Graphs and Multi-Agent Systems

3:34:07Generating Data and Building Knowledge Graphs

Generating Data and Building Knowledge Graphs

3:44:04Building and Deploying the Knowledge Graph

Building and Deploying the Knowledge Graph

Part 10: Open LLMs and Prompt Learning

3:53:56The State of Open LLMs in 2025

The State of Open LLMs in 2025

4:04:09Deployment Options and Lessons Learned

Deployment Options and Lessons Learned

4:11:11Trends in Open LLMs and Future Directions

Trends in Open LLMs and Future Directions

4:22:31Introduction to Prompt Learning

Introduction to Prompt Learning